Prediction Of Spot Price Of Iron Ore By Wavelet Transform And LSSVM Combined Model Based On The Theory Of Phase Space Reconstruction

With the promotion of the degree of economic globalization and the acceleration of the process of urbanization in China,the demand for iron ore and other commodities is expanding at the same time in China's industry.Due to the restriction of domestic resources and environment,China's dependence on imported iron ore is increasing day by day.In 2003,the import of iron ore in China has been the first in the world.Although China is the largest importer of iron ore in the world,it lacks influence in iron ore pricing.With the continuous evolution of the pricing method of iron ore,the price of iron ore has been more and more volatile,which has seriously affected China's iron ore production enterprises,consumer enterprises and national interests,so the forecast of iron ore prices has been widely concerned and studied.Aiming at the problem that the existing single time series model is not accurate and robust enough in forecasting of iron ore prices and the parameters of traditional LSSVM model is difficult to determine,a combined model based on wavelet transform and LSSVM(Wavelet-LSSVM)is proposed to predict the spot price of iron ore.In contrast simulation,ARIMA model,LSSVM model and Wavelet-LSSVM model were used to forecast the spot price data of 61.5%PB powder from June 1,2015 to June 8,2017 in Caofeidian port.The experimental results show that compared with the ARIMA model and the LSSVM model,the Wavelet-LSSVM combination model achieves better prediction results.At the same time,the model has a good performance in the multi-step prediction of iron ore price.